Sunday, 16 February 2014

Rationality Vs Its Literary Genre

I have long been somewhat skeptical of LessWrong. Recently, I read Holden Karnofsky's 2011 LW post explaining why he doesn't recommend the Singularity Institute to donors. The post is harsh and criticizes almost every aspect of SI but the point I found the most interesting point was the following:
"Many of the things that SI and its supporters and advocates say imply a belief that they have special insights into the nature of general rationality, and/or have superior general rationality, relative to the rest of the population. (Examples herehere and here). My understanding is that SI is in the process of spinning off a group dedicated to training people on how to have higher general rationality. Yet I'm not aware of any of what I consider compelling evidence that SI staff/supporters/advocates have any special insight into the nature of general rationality or that they have especially high general rationality...
"I endorse Eliezer Yudkowsky's statement, "Be careful … any time you find yourself defining the [rationalist] as someone other than the agent who is currently smiling from on top of a giant heap of utility." To me, the best evidence of superior general rationality (or of insight into it) would be objectively impressive achievements (successful commercial ventures, highly prestigious awards, clear innovations, etc.) and/or accumulation of wealth and power. As mentioned above, SI staff/supporters/advocates do not seem particularly impressive on these fronts, at least not as much as I would expect for people who have the sort of insight into rationality that makes it sensible for them to train others in it. I am open to other evidence that SI staff/supporters/advocates have superior general rationality, but I have not seen it."

Luke Muehlhauser only responded that rationality is a poor predictor of life success. But then how do we assess rationality?

One thing that can't be denied is that LessWrongers have a lot of information and ideas about rationality and how to be more rational in one's life. They're familiar with a lot of philosophy, cognitive science, and economics. They've come up with life hacks, mind hacks, productivity tricks like the Pomodoro technique, tools for reasoning, de-biasing techniques, etc. And these tricks are likely effective when practiced.

They're clearly high on epistemic rationality and they even have epistemic rationality about instrumental rationality, but do they have more instrumental rationality than ordinary smart people? I don't think there's evidence that knowledge (and perhaps use) of effective rationality techniques make people more instrumentally rational than others.

Take the following scenario. There are two humans, PlainJane and SuperNerd. PlainJane is a social, smart woman with no knowledge of philosophy, psychology, the literature on heuristics and biases, or any rationality techniques. She makes her decisions based on common sense, experience, and received advice. She is what Eliezer Yudkowsky would call a "muggle." SuperNerd is less social and extroverted but he studies the above topics on a daily basis. He knows how to apply Bayes Theorem, he practices his belief calibration, and so on.

Both PlainJane and SuperNerds want to accomplish the same goals. These could be to get the jobs they want, or to make friends, or to stay in a good mood, or to get healthier.

Who is more likely to accomplish these goals? Personally, I don't think one has a major advantage over the other. (If I'm stereotyping and drawing from information not provided in the character descriptions, then I'd bet on PlainJane to be better at accomplishing her goals.)

But let's assume both people managed to accomplish one of the goals, say to become more productive.

PlainJane thinks she needs to waste less time so she tries to avoid watching movies, declines to hang out with her friends one night so she can get work done, and concentrates harder whenever she's working.

SuperNerd looks into the literature on productivity and finds all sorts of hacks, tricks, and techniques for boosting productivity. Through careful practice, the instalment of new habits, and micro-optimizing his routine, he also manages to boost his productivity.

Who has demonstrated more rationality here?

I again feel as if there's no real difference between the two. They accomplished the same thing, snagged their utilons, and that's that. I think some people are biased toward SuperNerd in this case because he displayed more micro-rationality. He was jumping through hoops and de-biasing, while PlainJane doesn't even know roughly how brains work. But rationality isn't assessed by your proximity to the literary genre of rationality, it's assessed by your ability to make it rain utilons. PlainJane and SuperNerd both accomplished the same goal and both displayed the same proficiency at instrumentalizing their ends. Your chosen pathway to accomplishing the goal shouldn't be relevant.

Most of the rationality I've witnessed in my life has come from "muggles" like PlainJane rather than from people like SuperNerd. Ordinary people are decent at accomplishing their goals. Some knowledge, new techniques, and micro-adjusting could definitely improve on common sense, but my impression is that most rationalists are spending too much time obsessing over the literary genre of rationality rather than actually being rational and accomplishing things. If you consider yourself to be of high rationality but can't point to ways in which your life success is very high, than that should be a giant red flag. (One possibility is that you started off in such bad conditions that you had to be extremely rational just to get yourself up to an average place.) A good question to ask yourself is: Since "becoming more rational," how has my life changed? If it hasn't changed in any significant way, then you probably haven't become more rational.

Rationalists should put more effort on accomplishing concrete impressive things (awards, degrees, grants, starting companies, impressive jobs, displaying extra productivity, contributing to good causes) than to just talking about and testing out different rationality-genre subjects.

Personally, I have found it very interesting to learn about cognitive science, philosophy, and rationality but I'm not sure I've gained many utilons as a result. I'm Less Wrong but not necessarily More Rational. I would probably be More Rational if I spent less time trying to micro-optimize and more time just getting stuff done.


  1. Life hacks have always been the part of LW that interested me least. I don't really use life hacks myself, and when I have tried them, they've usually either not worked or led to worse consequences in the long run. I'm a very emotion-driven person, both naturally and because I've found this works well for me. I don't set explicit goals but instead just follow my passions and conscience where they lead me.

    That said, formal rationality may work for others, and in any event, it's an interesting science in its own right, just like other parts of psychology. But "rationality promotion" is not the reason I think MIRI has high value.

    LW rationality can also refer to something else: The naturalist worldview that cuts through philosophical paradoxes on free will, ethics, consciousness, etc. Here I think Eliezer's writings are exceptional, and it may not be an exaggeration to say that Eliezer ranks among the top 10 most influential people in my life through his writings alone.

    1. Yeah, I think my potential to have life success might have been improved by reading some of the Sequences and reducing my philosophical confusion. Learning about these things definitely changed my goals but I'm not sure they made me better at achieving my new goals. Most learning doesn't translate into tangible life or skill improvement.

      I really like the idea of self-improvement and life hacks but my impression is that most of it is about improving really narrow areas of one's life (e.g. sleeping pattern), possibly at the expense of other areas. I don't think measuring the success of a narrow life hack by how well it impacts that one area is a good way to determine whether it's a good life hack. You also need to look at your general life success before and after.

    2. Having a less confused worldview may be more important than any productivity boost because focusing on better things can multiply your impact many times.

      I think the best life hacks are taught in school or other mainstream venues (e.g., tips for success in business, etc.). Presumably the cutting-edge ones have lower returns in general (except maybe polyphasic sleep if you get it to work, but many people have tried and failed).

    3. Right, but now we're talking more about ethics (making life better for others) rather than rationality (making life better for yourself). I think we're on the same page here.